The efficacy of therapeutic granulocyte transfusions is limited by the relatively small number of cells obtained using standard steroid stimulation of the donor. To define an optimal mobilization schedule that maximizes cell yields while minimizing donor discomfort during granulocytapheresis, we have studied three donor mobilization schedules. Forty healthy donors have undergone three leukapheresis procedures each, receiving either dexamethasone (dexa) 8 mg orally 12 hrs prior to donation, granulocyte colony stimulating factor (G-CSF) 5 ug/kg SQ 16 to 24 hrs prior to apheresis, or dexa plus G-CSF (D+G) in the same doses. Seven liters of whole blood were processed on the CS-3000 Plus device using Hetastarch as the sedimenting agent. For C-CSF-mobilized procedures, a comparison was made between two interface offset (IO) detector settings on the CS-3000 device: an IO of 33 and an IO of 45.Administration of G-CSF alone led to a 3.8-fold increase. Use of dexamethasone plus C-CSF led to a 4.9-fold increase in donor peripheral blood granulocyte counts compared to dexa alone (from 6.3 ' 2.5 with dexa, to 24.1 ' 4.9 and 30.5 ' 6.2 x10 to the 9th power cells/L with G-CSF and D+G, respectively). Similarly, use of G-CSF alone or the combination of dexa and G-CSF resulted in 2.3- and 3.5-fold increases in granulocyte content in the product compared to dexa alone (from 2.09 ' 0.68 with dexa alone to 4.87 ' 1.02 and 7.31 ' 1.56 x10 to the 10th power cells total with G-CSF and D+G, respectively), p is less than .01 for all comparisons between dexa and either G-CSF or D+G. In addition, granulocyte counts in both the donor's blood and the apheresis product were greater when donors took the combination of D+G vs G-CSF alone (27 percent increase in the blood granulocyte count and 49 percent increase in product granulocyte content when dexa was added to G-CSF) (p is less than .05). Increasing the IO of the device from 33 to 45 did not increase the efficiency of the procedures, and we are continuing to make modifications to the run parameters of the CS-3000 machine to optimize the mechanics of granulocyte harvesting. With dexa alone, 44 percent of donors had insomnia or flushing; with G-CSF, 68 percent had bone pain, headache, insomnia, or fatigue; this increased to 72 percent with combination D+G. Ten of 100 donors requested discontinuation of G-CSF-mobilized donations due to discomfort or inconvenience.Addition of dexa to G-CSF significantly increases granulocyte yields and is accompanied by a modest but well-tolerated increase in donor inconvenience and discomfort. The availability of granulocyte products with markedly increased numbers of cells is resulting in a renewal of interest in this transfusion component.